A Realist's View on the Job Market

More than one in three Americans older than 16 is not working. More than 88 million people, about 36.5% of that age group, are without jobs, according to the US Dept. of Labor's Bureau of Labor Statistics (BLS). The light blue line in the chart below is a plot of the jobless going back to 1990, and the current number is bigger than at any time since February 1984, when we were recovering from the stagflation of the late 1970s.

Percentage of US Civilians Not Working

Year-over-year change through April 2012

Yet the Dept. of Labor boasts every week that millions of jobs have been created in the past few years. How is it possible for them to report millions of new jobs and a declining unemployment rate when there are more people without jobs than at any time since 1984?

Here's how the government does it. Look at the chart again. The black line represents the percent of jobless people among those the government qualifies as members of the labor force -- people older than age 16 who admit that they have recently looked for a job. The number is falling because more and more people have stopped looking for work. Either they have returned to school, are living with their parents, or have simply given up looking for work.

The next chart -- showing reported and actual average change in initial jobless claims -- was last posted on this Website at the end of December to illustrate that the Dept. of Labor revised its weekly figures upward by an average of 3,286 newly unemployed people in 56 of the previous 58 weeks.

Twenty-one weeks have now passed since then, and 20 more upward revisions have been made. Every week, like clockwork, the DOL discovered that the number of newly unemployed reported the previous week was actually, on average 3,628 higher than reported, 10% worse than was reported in December. One would think that with all this experience they might be able to reduce the number and size of their errors. But no, the DOL statisticians have increased them.

Using this trick, they were able to report an improvement of 1,000 last week when the reality was actually 2,000 worse than reported the week before. Note that the slope of the revisions has become steeper since March, probably because the employment situation is getting worse, so there is more to hide.

Reported and Actual Average Change in Initial Jobless Claims

Through April 2012

It's clear that the three-month moving average of newly unemployed has rolled over and is climbing again (note that we have inverted the right axis to show a reduction in unemployment as a positive event).

It is also clear that the stock market is doing what it always does when the employment situation worsens -- it falls, as it has for 12 years. (The correlation coefficient between the stock market and the level of unemployment is -69%, which tells us that when unemployment increases the stock market falls 69% of the time.)

Initial Jobless Claims -- Seasonally Adjusted

Through Thursday, May 24th

Instead of focusing on the number of people who have been searching for work, why not focus on the 58.5% of Americans who actually have jobs? The BLS divides the number of working Americans by the number of people living here and reports this figure each month.

Using this information, we can finally get an honest count. The St. Louis Federal Reserve posts the statistics on its Website for all to see. In addition, the official version of the next chart goes all the way back to 1947. It is clear from the data that we are still near the bottom, at a level not seen since 1984.

Percentage of Civilians Who Are Working

Through April 2012

Since the government has made accurate figures available to those who know where to look, why does it emphasize figures that do not accurately reflect the employment situation? My guess is that in an election year, the incumbent benefits from a rising stock market, and the stock market trends higher when people are working.

I previously posted this chart of the 30-day average of presidential popularity with potential voters according to the Rasmussen daily poll. The chart is based on the net difference between the number of voters who strongly approve of Barack Obama and those who strongly disapprove.

Rasumssen Poll -- Obama Popularity

Through through Monday, March 12th

The correlation between the SP 500 and Obama's popularity has been 61% since November 2010. In other words, in the past 18 months, when the market advanced, Obama's popularity increased 61% of the time, according to the Rasmussen poll. Perhaps this explains why the employment figures are presented as they are.

— Fred Goodman, a registered investment adviser and Certified Financial Planner, publishes MarketMonograph, a daily Web-based subscription service specializing in technical stock market analysis and the application of economic indicators to market timing. You can reach him at fred@marketmonograph.com.